This invention relates to communication methods and apparatus for providing network management, bandwidth and path control in a heterogeneous network that may be composed of multiple vendor equipment and transmission paths. More specifically, the communication system concerns semi-autonomous implementation components within a management hierarchy to globally manage multiple vendor elements while satisfying local network demands.
Telecommunications services have, for many years, attempted to optimize or minimize bandwidth usage between network elements. Since the modern communications era, brought about by the theories of Shannon, telecommunications engineers have been keenly aware of the need to provide optimal, or at least good solutions, to bandwidth allocation problems in point-to-point and point-to-multipoint networks.
In wireless communication systems, solutions to bandwidth allocation problems can be seen in the way data is modulated to xe2x80x9csharexe2x80x9d finite resources. For example, time division multiple access (xe2x80x9cTDMAxe2x80x9d) provides a means for multiple stations to access time slots on satellite carriers and thereby xe2x80x9csharexe2x80x9d bandwidth resources. Code Division Multiple Access (xe2x80x9cCDMAxe2x80x9d) provides a means to use code division modulation techniques (time and frequency modulation) for multiple point access to a predetermined range of bandwidth and thereby xe2x80x9csharexe2x80x9d bandwidth space. Likewise, frequency division multi-access (xe2x80x9cFDMAxe2x80x9d) provides a means to divide up and share a finite bandwidth resource.
More elaborate schemes to dedicate bandwidth in accordance with a predetermined transmission schedule and modulation plan can be seen in U.S. Pat. No. 5,592,470 to Rudrapatna et al., (xe2x80x9cRudrapatnaxe2x80x9d) issued Jan. 7, 1997, (the xe2x80x9cRudrapatna patentxe2x80x9d). The Rudrapatna patent concerns a terrestrial micro-port network that allocates bandwidth to terrestrial micro-port receivers based on a pre-determined schedule and modulation plan. The pre-determined schedule and plan may be subsequently modified by dynamic demands on the micro-ports. The network can then satisfy the dynamic demands by moving channels between modulation and polarity schemes in pre-determined amounts.
In wireless networks, certain communications links require more bandwidth and power resources than others. This is necessary to maintain specified information throughput, to provide appropriate grades of service or due to varying site configurations (e.g., different antenna sizes). Whenever a change in network resource allocations is required to match varying traffic requirements, a new transmission plan may or may not be implemented. This may necessitate programming, transmitting and receiving communications equipment, e.g., amplifiers, modulators and demodulators, to support the new resource assignments. These and other problems in bandwidth allocation in a multi-vendor network are addressed by the present invention.
The methods and apparatus disclosed herein may assign and re-assign available transmission resources in point-to-point, multipoint and broadcast wireless networks. This may be accomplished on the basis of information capacity and connectivity requirements between transmitters and receivers of communications links at the time of assignment or reassignment. The system may also provide a network administrator with novel tools and automated methods to define and implement network transmission plans and to modify allocation decisions as traffic requirements change.
The system may provide the tools to efficiently allocate transmission resources. These tools help implement the communications links that form wireless networks. An optimum resource or a xe2x80x9cgood fitxe2x80x9d allocation is achieved when network users have just enough information transmission capacity to perform their tasks. One way to accomplish optimal or good transmission resource allocations in a wireless network is to analyze network users"" usage patterns and allocate capacity according to a time-varying schedule.
By analyzing network usage patterns, a management component can determine a transmission plan schedule that efficiently allocates the satellite bandwidth available to the network based on historical usage patterns. The management component may automatically schedule and implement a transmission plan. As the network users"" requirements change, the management component may update or modify the scheduled transmission plans to satisfy the new requirements.
The system may automate implementation of transmission plans by reprogramming the system when predetermined parameters are reached. For example, the management component may determine a transmission plan from a historical analysis of bandwidth requirements between stations. This transmission plan may be automatically deployed to the network. The management component can then monitor and analyze network allocation demands to determine a new transmission plan. The new transmission plan can then be automatically deployed in the network when predetermined parameters are reached, such as, average change in bandwidth, e.g., bandwidth in use/bandwidth in the transmission plan, exceeds a predetermined amount or if a predetermined amount of time has transpired. The transmission plans may be propagated as generic network commands and translated into corresponding equipment parameters and associated control commands as required for reconfiguring network equipment elements. Thus, the system may generate and distribute equipment configurations to network elements to reprogram for synchronized execution at predetermined times.
The system further controls and schedules bandwidth between network elements to consider other network factors such as economic constraints. In a wireless communications network, each communications carrier should have just enough bandwidth and power necessary to meet the needs of its corresponding users. Although optimum resource allocation is the primary goal, sub-optimum allocation may be tolerated when economic constraints may limit transmission resources to finite amounts. Thus, for example, a dynamic bandwidth requirement at a network station may require an increase in bandwidth allocation from the station, such as when the queuing depth reaches a predetermined amount at the station switch. The station may have additional capacity available on an available communication link, however, the incremental capacity of the link may far exceed the bandwidth required to reduce the depth of the communication queue. Furthermore, the financial cost of the incremental capacity may exceed the cost of waiting for network usage to decrease to reduce the depth of the queue. The system, in this case, would allow the network to back up and flow control the user data before the system would allocate additional capacity. The system provides methods to use finite transmission resources by enabling power and bandwidth to be re-allocated as needed to meet changing communications requirements in satellite networks. However, the capabilities of the system are applicable to all wireless networks that can be modeled as a collection of transmitters, transmission resources, and receivers.
The system provides a means to manage heterogeneous or multiple vendor network equipment over heterogeneous or multiple vendor transmission resources with multiple transmission paths. One such path may be via programmable C-, Ku-, or Ka-band satellite networks. Other paths may be via discrete carriers available on a preprogrammed networks such as the Inmarsat, Globalstar or Iridium satellite systems. Yet other paths may be via third party medium or broadband networks such as the envisioned Teledesic satellite network. Yet another path may be over a programmable or managed network such as the Intelsat global satellite system. Thus, the system provides a means to define and manage capacity between network elements where the network may be a combination of a discrete bandwidth allocation network managed by an external system, a semi-programmable medium or broadband network wherein a varying amount of bandwidth may be allocated from an externally managed resource and a fully-programmable network where the resource is managed by a network management component. Thus, the management system provides a nearly transparent means by which an operator, user or network element may place demands on the network and the management system may satisfy those demands based on a least cost algorithm, quality of service parameters and pre-defined or time-varying transmission plans.
The management system described may configure the transmission elements (transmitters and receivers) in a wireless network to implement a specified allocation of transmission resources according to varying, scheduled or ad-hoc capacity requirements. The system maintains a schedule of transmission plan implementations and may automatically perform each implementation at a scheduled time.
The semi-autonomous network management system essentially consists of two semiautonomous components. The first component is the Implementation Component (IC) which executes at a site containing network transmission elements and the second is a Management Component (MC) which executes at a network management site. These components may be connected via a user datagram Internet protocol messaging format.
At the heart of the system is the IC. The IC may be a stand-alone application program that controls one or more network elements. A network element may be the station or communication equipment physically controlled by an IC. Thus, it is usually the case that the network element is a stationary or mobile communications node at a single physical location. The IC may, however, remotely control a network element.
In one embodiment of the present invention, the IC application may execute in a dedicated processing environment such as a PC executing UNIX or other suitable operating system such as Windows or DOS. The IC may also execute on an integrated or embedded processor such as a PC on a card environment with an application specific or real time operating system executing thereon.
The IC is semi-autonomous, e.g., it can translate allocation commands from a management component into executable commands for its associated network elements without having direct, full-time contact with the network management component. The IC may store pre-programmed parameters, transmission plans, or collection commands for automatic execution. The IC may map a network programming language command set or generic allocation command to a vendor specific command sequence. The IC may contact the management component to receive permission to access network bandwidth, to report the unavailability of network elements, or to request different allocation parameters if or when local demands exceed the IC""s preprogrammed allocations. Thus, the IC may provide independent control over network elements while maintaining or executing a management plan.
In the semi-autonomous network management scheme disclosed, transmission schedules may be loaded in advance of the implementation of the scheduled transmission plans. Then, at a predetermined time, the network can switch over to the new transmission plan to implement the optimal, or at least good solution, before more complicated dynamic bandwidth allocation algorithms would need to be employed.
In addition to automatically implementing scheduled transmission plans generated by the management component, the system may also perform network usage analysis. Automated network usage analysis may require that the management component have access to traffic data collected for the network. The data may be collected automatically or manually by the management component or the implementation component may interact with the elements in the network to collect the usage data. The management component may use statistical methods to analyze the gathered network usage data to suggest or implement optimize transmission plans for efficient use of the available resources according to a schedule.
Efficient use of bandwidth spectrum may be achieved on various levels in the system. On a first level, bandwidth may be scheduled in accordance with a historical analysis of demands on the network. For example, it may be determined that Monday morning traffic is mostly outbound (i.e., from a central earth station to a mobile station). On Fridays, however, most of the traffic is in the opposite direction (i.e., from mobile stations back to the central earth station). In this instance, an assymetric channel may be opened for Monday traffic to provide higher outbound data and a slower speed return path. Then the opposite allocation may be established for Friday""s traffic (e.g., a high speed channel from a mobile station to the central station and a low speed acknowledgment channel from the central station back to the mobile station). This may provide an optimal, or at least a cost-effective solution for the capacity requirements at a particular time.
On a second level, the system may allocate capacity based on class of service parameters available, for example, through an Asynchronous Transfer Mode (xe2x80x9cATMxe2x80x9d) type packet format. For example, a class of service may identify data packets with low priority for a particular application. In such a case, an expensive satellite carrier may not be necessary and a lower-cost transmission resource may be put online by the network to pass the required data packets. Thus, the present network can mix class of service bandwidth allocation methods with least cost routing decisions based on predetermined parameters programmed in the IC.
The semi-autonomous nature of the network management components may use a datagram protocol for interprocess communication. More specifically, the network components may communicate through the use of the User Datagram Protocol (xe2x80x9cUDPxe2x80x9d) over an internet protocol (xe2x80x9cIPxe2x80x9d). Communication between the management component and the IC may use a polled or interrupt methodology.
In a polled mode, the management component contacts each of the ICs to pass UDP/IP messages or to receive Transmission Control Protocol/Internet Protocol (xe2x80x9cTCP/IPxe2x80x9d) information from the particular IC. In an interrupt driven mode, the IC may attempt to communicate with the management component. The interrupt mode may be used to reestablish contact with the MC if the IC loses synchronization with the network or to pass alarm or other local conditions happening at the IC that may not be detected by the management component. In the interrupt driven mode, the IC may have a preassigned backup channel or predetermined bandwidth allocation to communicate with the management component. The management component may be programmed to look for alarm conditions or communication attempts from the ICs when predetermined parameter thresholds are reached.
Additionally, a signaling control mechanism between the management component and the ICs is disclosed. The signaling control mechanize operates to ensure that each of the ICs receives the appropriate message(s) and that the transmission plan may be implemented according to the predetermined schedule.
The signaling control mechanism between management component and implementation components may communicate by exchanging the following UDP/IP messages.
Transmission Control Order
Abort Order
Acknowledgment of a Transmission Control or Abort Order
Audit Request
Audit Response
A transmission control order (xe2x80x9cTCOxe2x80x9d) specifies new transmission parameters for a transmitter or receiver. The TCO may also specify the implementation time. The implementation time may be the time at which elements should begin using the transmission parameters specified in the order. TCOs are generated by the system to implement a new transmission plan. The system sends TCOs to the ICs of transmitters and receivers that must change parameters to implement the new transmission plan. TCOs may be stored on a hard drive or other non-volatile storage so that they are preserved through IC restarts and at IC power failures.
It is possible that an IC may be down or may not be able to communicate with the managed equipment at the execution time of a TCO. When this happens, the IC may implement the current transmission plan or may implement a default state when the IC reestablishes communication with the managed equipment.
The IC may send an acknowledgment of a TCO when a TCO is received from the system. If any of the requested parameter changes cannot be implemented because the managed equipment or the configuration files do not support it, the IC notifies the management component of this in the acknowledgment.
The IC may also check that the parameter values are valid for the managed network equipment. Parameter ranges are specified in the Equipment Controller configuration files in the IC, discussed further below.
A confirmation message may not be necessary for reporting the successful implementation of a transmission control order. Because the majority of satellite networks implement single links to each remote site, if an IC is not able to implement a TCO, the IP connection to the system may be lost. The system management component may detect the problem from the lack of audit responses. If the system does not receive an audit response from an IC, the system may update the site status and alerts the management component alarm.
An abort order may instruct the IC to cancel any pending TCO for the specified transmitter or receiver. The system may send abort orders when a pending implementation is canceled by the Administrator. The IC may send an acknowledgment when it receives an abort order. The IC may send an acknowledgment when it receives a TCO or abort order from the management component.
An audit request may be periodically sent to an IC by the management component. The management component may send an audit request to check the status of a transmitter or receiver. One audit request may be sent for each transmitter and receiver being managed by an IC.
An audit response may be sent by an IC when an audit request is received from the management component. The audit response may contain the current parameter values for the transmitter or receiver specified in the audit request.
An audit response may be similar in structure to a TCO. It may include the hardware identification for a transmitter or receiver and a list of model parameters and their current values as reported by the physical hardware.
The receive frequency model parameter may be a special case: the frequency reported by the demodulator may not match the commanded receive frequency. Sources of frequency error throughout a wireless carrier transmission process may result in an offset between the expected and actual receive frequencies. Many demodulators are designed to accommodate this frequency offset by searching for the transmitted carrier in a neighborhood around the commanded receive frequency. However, the system may also account for this receive frequency offset when determining whether the physical hardware is using the same parameters as in the most recently implemented TCO.
The management component may periodically request the current parameters from all transmitters and receivers. This network auditing function may perform the following functions:
Maintains the status of communications between the management component and the transmitters and receivers in the network.
Detects parameter changes of the managed equipment.
When a difference between the specified transmission parameters for a transmitter or receiver and the managed equipment is detected, the management component may notify a Bandwidth Administrator. The management component operator interface may use an audible as well as visual alert to improve the chance that a Bandwidth Administrator will notice the difference and act to resolve it.
FIG. 21 shows equipment controller IC (150). IC (150) may have a configuration database (152) which stores a configuration mapping for end-user receiver and transmitter equipment which is interfaced by, for example, serial devices (164, 166, 172, 174) and parallel devices (188, 190). The receiver/transmitter equipment may be from multiple vendors and thus the configuration database (152) maps commands from the management component (156) to a particular device. This feature of the IC may allow the use of a generic network control language in commands sent to IC (150) (discussed further below).
The system is designed to manage all transmission equipment, regardless of manufacturer. To achieve this, the management component deals with model satellite transmitters and receivers as illustrated in FIG. 6.
The transmitter and receiver models have the parameters necessary to implement a wireless link. Only parameters that relate to the establishment of a wireless link need be included in the transmitter and receiver models.
The management component may not require information about the physical equipment elements used to implement the communications links in the managed network. Therefore, the MC need not map the model parameters directly to commands for the physical hardware of the transmitters and receivers at a site. The IC may have information about the physical hardware at its site and may map the model parameters to the appropriate commands and responses.
The IC may read information about the physical hardware from the configuration files. These files may specify the information required by the IC to monitor and control the managed equipment at a site. The IC configuration files may contain the information necessary to convert parameter changes for the model transmitters and receivers into commands to the physical hardware.